HYDROCARBONS AND BTEX PICKUP AND CONTROL FROM AMINE SYSTEMS Jerry A. Bullin William G. Brown Bryan Research & Engineering, Inc. Bryan, Texas, U.S.A. ABSTRACT HC and BTEX absorption into amine solutions has received increased attention over the last decade due to emissions to the atmosphere or to problems in downstream equipment. The collection of VLE and VLLE data by GPA and others have facilitated the development of a model for the absorption and removal processes. The amount of HC and BTEX emitted or passed to downstream equipment may be controlled by reducing the absorption or by removal from the rich amine. 1
HYDROCARBONS AND BTEX PICKUP AND CONTROL FROM AMINE SYSTEMS INTRODUCTION Over the last decade, the problems created by the absorption of hydrocarbons (HC) and aromatic compounds consisting of benzene, toluene, ethylbenzene and xylenes (BTEX) in various areas of gas processing have become widely recognized. The problems include HC and BTEX emissions to the atmosphere as well as to further downstream processes. The two primary areas for BTEX absorption occur in amine sweetening and glycol dehydration. In many cases, the amine sweetening unit is followed by a glycol dehydration unit. In a typical amine sweetening unit as shown in Figure 1, acid gases (consisting mainly of H2S and CO2) along with some HC and BTEX (if present in feed gas) are absorbed by the amine solution in the contactor. Depending upon the pressure in the contactor, a flash tank immediately downstream of the contactor allows some of the acid gases, HC and BTEX to be flashed from the amine solution. Nearly all of the remaining acid gases, HC and BTEX are removed from the amine solution in the regenerator. Figure 1 Simple Amine Sweetening Facility If the flash tank and regenerator are vented, the acid gases, HC and BTEX would be directly emitted to the atmosphere. The Clean Air Act limits the amounts of heavy hydrocarbons (Volatile Organic Compounds) which may be emitted from a facility to 25 tons per year. Aromatic compounds such as BTEX are limited to a cumulative amount of 25 tons per year and 1 tons per year of any individual aromatic. 2
If the acid gases are fed to a sulfur recovery unit, the heavy HC and, particularly, BTEX are known to cause problems. BTEX are difficult to burn and can cause coking in the catalyst beds. In split flow plants where part of the acid gas stream is passed around the burner, some of the BTEX can pass through the catalyst beds and into the tail gas. Thus, to adequately investigate HC and BTEX pickup and control, it has become increasingly important to develop accurate and reliable methods to predict the absorption, desorption and control of HC and BTEX in amine sweetening units. For the development of the methods, vapor-liquid equilibria (VLE), vapor-liquid-liquid equilibria (VLLE) and thermodynamic data must be available along with plant operating data for verification. As reviewed below, this data has, historically, been very limited. VLE AND OPERATING DATA McIntyre et al. [1] presented an excellent review and summary of the VLE, VLLE and operating data for HC and BTEX in amine systems. Their review is summarized below and extended to cover other sources of data as well as the recently available Project 971 [GPA Research Report 971 (RR-971)] by Valtz et al. [2] and to include the impact of acid gas loading on the solubility of HC and BTEX in amine solutions. Since increased acid gas loading can reduce the HC and BTEX solubility by as much as 3 to 4% (RR-971), it is an essential feature in any analysis. VLE Data The published VLE data are summarized in Table 1. As can be seen from this table, prior to the data by Critchfield et al. [13] in 21, the HC data were limited to methane and ethane except for the propane and n-butane in MDEA data by Carroll, et al. [5] and Jou et al. [6]. Furthermore, the propane and n-butane in MDEA data did not include the impact of acid gas loading. As can also be seen from Table 1, prior to the data by Hegarty and Hawthorne [11] in 1999, and the RR-971 [2] data in 22, there was almost a total void of data for BTEX. From the BTEX data, the only data including acid gas loading was for toluene with CO2 (RR-971). Operating Plant Data As discussed by McIntyre et al. [1], there are only about 1 cases reported in the literature where some absorption data for HC and BTEX are measured from operating plants. From these, only about 4 or 5 give nearly complete data. The sources of the available plant operating data are presented in Table 2. Further details on the operating data can be obtained from McIntyre et al. [1] or from the original work cited in Table 2. MODEL DEVELOPMENT Along with the data for solubility of the various HC and BTEX in water, the above referenced data were used to develop a model for the solubility of HC and BTEX in amine solutions. This model has been incorporated in TSWEET and is currently being incorporated into ProMax with TSWEET and PROSIM. The current TSWEET model calculates the HC and BTEX solubility as a function of temperature, pressure, amine type and amine concentration. The TSWEET model does not include the effect of acid gas loading on solubility and is based on a limited portion of the RR-971 data. The ProMax model will include the effect of acid gas loading and will incorporate all of the available data. The TSWEET model was used in this work to perform parametric studies and illustrate trends except as noted in the discussion. 3
Table 1 Sources of Solubility Data for HC and BTEX in Amine Solutions Year Authors Solvent Concentration Component Temperature 1976 Lawson and Garst [3] a MEA, DEA 5, 15, 25, 4 wt% Methane, Ethane Acid Gas Loading 1-25 F - 1986 Dingman [4] DGA 5 wt% Methane 15-19 F - 1992 Carroll et al. [5] MDEA 3 M Propane 25-15 C - 1996 Jou et al. [6] MDEA 3 M n-butane 25-125 C - 1996 Jou et al. [7] TEA 2, 3, 5 M Ethane 25-15 C - 1998 Jou et al. [8] DGA 3, 6 M Methane 25-125 C - 1998 Jou et al. [9] MDEA 3 M 1998 1999 21 21 22 Carroll et al. [1] Hegarty and Hawthorne [11] Addicks, et al. [12] Critchfield et al.[13] Jou et al. [14] 22 Valtz et al.[2] 22 Addicks and Owren [15] MEA, DEA, TEA MDEA DGA Methane, Ethane 25-13 C - 3 M Methane 25-125 C - 25, 5 wt% 35 wt% Benzene, Toluene Benzene, Toluene 25-12 C - 6 C - MDEA Methane 4, 8 C Limited MEA, DEA, DGA, MDEA, DIPA MDEA, DGA, DEA, MEA 3, 4.5 M Varying a Addendum in 1996 by Mather and Marsh [23]. Methane, Ethane, Propane, n-butane, n-pentane, Benzene Benzene, Toluene, Ethylbenzene, Xylene 4 C - Varying Limited MDEA 3 5 wt% Methane 4-8 C Yes 4
Table 2 Sources of Plant Operating Data for HC and BTEX in Amine Sweetening Units Year Authors Amine Number of Plants with Data 1972 Harbison and Dingman [16] DGA 1 1981 Huval and Van de Venne [17] DGA 1 1984 Law and Seidlitz [18] 1 1996 Morrow [19] - 1 1997 Skinner et al. [2] DEA MEA 1999 Hegarty and Hawthorne [11] MDEA 1 5 1 METHODS TO CONTROL HC AND BTEX ABSORPTION AND EMISSIONS The methods to control HC and BTEX absorption and emissions may be divided into two types. The first is the selection of amine process conditions to reduce absorption. The second is the use of additional equipment to capture the absorbed HC and BTEX and prevent their emission into the acid gases. Reducing HC and BTEX Absorption Although in most cases many of the amine process conditions will not be flexible, full advantage should be taken of those that can be adjusted. In some cases, the process conditions to reduce the HC and BTEX absorption will conflict with the primary objective of acid gas removal and thus are not feasible. Since essentially all of the HC and BTEX absorbed in the contactor will be removed from the solution in the flash drum and regenerator, the emissions will be essentially equal to the amount absorbed. Again, the primary process conditions affecting absorption of HC and BTEX are temperature, pressure, amine solution circulation rate and amine type, concentration and acid gas loading. The impact of these process conditions were explored by parametric studies using the gas analysis shown in Table 3 except as noted below. In general, the other process parameters were changed to demonstrate the impact of the parameter under discussion. For example, the inlet gas temperature was manipulated to yield a range of rich amine temperatures. Temperature The effect of temperature on the absorption of methane, propane and benzene are shown in Figures 2 and 3. Since methane is above its critical temperature, its absorption is not impacted as much by temperature as the other HC and BTEX. The large amount of methane absorption compared to the other HC and BTEX is due to its very high partial pressure relative to the others. As can be seen from these figures, the absorption of the propane and benzene decreases by 3 to 4% with increased temperature from 9 to 14 F. Unfortunately, this same trend applies with acid gases except for some cases where CO2 absorption is kinetically limited. However, in certain situations, it may be possible 5
to increase the absorber temperature and still meet the sweet gas specifications. For treating of LPG liquids, the absorption would behave in the opposite manner to gases and would increase with rich amine temperature because the solubility of liquids increases with temperature. Table 3 - Gas Stream Used for Parametric Studies Composition Mole % C 1 83 C 2 6 C 3 4 C 4 2 C 5.5 C 6.38 CO 2 4 Benzene.6 Toluene.4 Ethylbenzene.1 Xylene.1 T = 85 F P = 1 psia Flow = 3 MMSCFD.2 HC & BTEX Absorption in Rich Amine (SCF/Gal).16.12.8.4 Methane Propane Benzene 9 1 11 12 13 14 15 Rich Amine Temperature ( o F) Figure 2 - The Influence of Rich Amine Temperature on HC & BTEX Absorption in 5 wt% MDEA 6
HC & BTEX Absorption in Rich Amine (SCF/Gal).25.2.15.1.5 Propane Benzene 9 1 11 12 13 14 15 Temperature o F Figure 3 - The Influence of Rich Amine Temperature on HC & BTEX Absorption in 5 wt% MDEA Pressure The absorber pressure directly affects the partial pressure of the HC and BTEX and thus the amount absorbed. In almost all cases, the pressure is dictated by other constraints with very little flexibility. For treating of LPG liquids, pressure has little effect on absorption. Amine Solution Circulation Rate Since a relatively small portion of the HC and BTEX are absorbed by the amine solution, the amount of HC and BTEX picked up increases directly with circulation rate for any given amine type and concentration as shown in Figures 4 and 5. Thus, the circulation rate should be maintained as low as possible but not exceed the maximum acid gas loading that would lead to excessive corrosion. Amine Type and Concentration The RR-971 data for VLLE systems can be used to illustrate the relative solubility of HC and BTEX in various amine solutions as shown in Figure 6 for toluene. In regard to the large solubility differences between the two DGA and two MDEA concentrations, Valtz et al. [2] have suggested that the solubility is a near logarithmic function of the amine concentration. 7
12 Total Pickup in Rich Amine (SCFH) 1 8 6 4 2 Methane Propane Benzene 6 7 8 9 1 Rich Amine Flow (GPM) Figure 4 - The Influence of Circulation Rate on Total HC & BTEX Pickup for 5 wt% MDEA 12 Total Pickup in Rich Amine (SCFH) 1 8 6 4 2 Propane Benzene 6 7 8 9 1 Rich Amine Flow (GPM) Figure 5 - The Influence of Circulation Rate on Total HC & BTEX Pickup for 5 wt% MDEA 8
.25.2 Solubility (SCF/gal).15.1.5 14.6 wt % MEA 35 wt % DEA 35 wt % DGA 52.5 wt % DGA 25 wt % MDEA 5 wt % MDEA 4.8 Mol % 8.5 Mol % 8.5 Mol % 15.7 Mol % 4.8 Mol % 13.2 Mol % Figure 6 - Relative VLLE Solubility of Toluene in Amine Solutions at 1 F (Data from RR-971 [2]) Using the gas stream described in Table 3, the methane, propane and benzene absorption in the rich amine is compared in Figures 7-9 for water, 25 and 5 wt% MDEA at various rich amine temperatures. Figure 7 shows that the amine has very little impact on the absorption of methane. For propane, Figure 8 shows about a 1% increase in absorption from water to 25 wt% MDEA and about a 2+% increase from 25 to 5 wt% MDEA. For benzene, Figure 9 shows about a 1% increase in absorption from water to 25 wt% MDEA and nearly a 1% increase again from 25 to 5 wt% MDEA. Unfortunately, these are, again, the same trends that favor acid gas pickup. However, it may be possible to select both the amine and the concentration and still meet the sweetening requirements. Amine Loading As illustrated in Figure 1 for benzene, the solubility of the HC and BTEX may be reduced by as much as 3 to 4% at higher acid gas loadings. This observation has been verified with the data Critchfield et al. [11] and RR-971 [2] for VLLE systems. Obviously, the acid gas loading must be maintained sufficiently low to prevent excessive corrosion. Higher acid gas loadings have the added benefit of reducing the circulation rate which, in turn, reduces the HC and BTEX absorption. Thus, reducing the circulation rate to yield the higher acid gas loadings has a double benefit on reducing HC and BTEX absorption. Overall Observations As previously stated, the above analyses were performed to isolate and illustrate the impact of a single process parameter upon the absorption of HC and BTEX in the rich amine. In any plant application, many parameters will change at once and simulations should be performed to determine the solvent and process conditions that will minimize the total HC and BTEX absorption. In general, the total absorption from a gas into the rich amine can be reduced by higher rich amine temperatures, lower column pressures, lower amine circulation rates, higher acid gas loadings, lower amine concentrations and a lower molecular weight amine. 9
Methane Absorption (SCF/Gal).25.2.15.1.5 5 wt % 25 wt % Water 8 9 1 11 12 13 14 15 Rich Amine Temperature ( o F) Figure 7 - Methane Absorption in MDEA Solutions Propane Absorption (SCF/Gal).16.12.8.4 5 wt % 25 wt % Water 8 9 1 11 12 13 14 15 Rich Amine Temperature ( o F) Figure 8 Propane Absorption in MDEA Solutions 1
.25 Benzene Absorption (SCF/Gal).2.15.1.5 5 wt % 25 wt % Water 8 9 1 11 12 13 14 15 Rich Amine Temperature ( o F) Figure 9 Benzene Absorption in MDEA Solutions.25.2 Benzene Solubility (SCF/gal).15.1.5.1.2.3.4.5.6.7.8.9 1 CO2 Loading Mole/Mole Figure 1 Solubility of Benzene in 5 wt% MDEA-Water-CO 2 Mixture at 77 F (Data from RR-971 [2]) 11
Removing HC and BTEX from Rich Amine If the HC and BTEX absorption into the amine solution cannot be reduced sufficiently to meet the emission limits then other control methods must be used. These methods include a hot flash and a stripping column to remove HC and BTEX from the rich amine. Hot Flash As can be seen from the schematic of the hot flash method shown in Figure 11, a flash tank can be added in the rich amine stream between the lean/rich exchanger and the regenerator. If the H2S content of the flash gas is above the site limits, a small amine absorber may be added onto the flash tank. The flash gas may be used as fuel thus destroying the HC and BTEX. Figure 11 Amine Sweetening with Hot Flash To illustrate the performance of a hot flash, an amine system using 5wt% MDEA with the feed gas shown in Table 3 was simulated. As shown in Figure 12, about 8% of the methane and propane would be flashed at 15 F. However, for benzene, significant flashing starts at 15 F. This is most likely associated with and largely resulting from the flashing of the CO2. It should be noted that, due to the relatively large amounts of CO2 in the rich amine, the flashing of even 1% of the CO2 produces very large volumes compared to the benzene. For this example, a hot flash would not be effective due to the large volumes of CO2 that would also be liberated. 12
12 C1 1 8 C3 Benzene CO2 % Flashed 6 4 2 1 14 18 22 Flash Temperature ( o F) Figure 12 The Influence of Temperature on HC & BTEX Vaporization from a Heated Flash of the Rich Amine HC and BTEX Stripping Column In this control method, a stripping column is added in the rich amine as shown in Figure 13. A portion of the sweet gas is used to strip the HC and BTEX from the amine solution. As in the hot flash, if the H2S content in the overhead stripping gas is above the site requirements, then an acid gas absorption section must be added to the column. This method has been patented and is described in the literature (McIntyre et al. [1], Morrow [19], Wallender and Morrow [21] and Morrow and Lunsford [22]). HC and BTEX removal greater than 75% are possible with the method (Wallender and Morrow [21]). The performance of an example HC and BTEX stripping column is shown in Figure 14 for 5 wt% MDEA treating the gas stream shown in Table 3. As can be seen from Figure 14, about 7% of the benzene can be removed from the rich amine at 12 F using about 5 SCF of stripping gas per gallon of solution. At these conditions about 1% of the CO2 would also be removed. The propane removed increases rapidly up to about 6% with the first introduction of the stripping gas. The drop in methane and propane removal is due to cooling in the stripping column and reabsorption from the stripping gas. 13
Figure 13- Amine Sweetening with HC & BTEX Stripping Column 8 7 6 % Removed 5 4 3 2 C1 Benzene C3 CO2 1 1 2 3 4 5 6 Stripping Gas (SCF/gal) Figure 14 Performance of a HC & BTEX Stripping Column on a Rich Amine with 5 wt% MDEA at 12 14
SUMMARY AND CONCLUSION HC and BTEX absorption into amine solutions has received increased attention over the last decade due to emissions to the atmosphere or to problems in downstream equipment. The collection of VLE and VLLE data by GPA and others have facilitated the development of a model for the absorption and removal processes. The amount of HC and BTEX emitted or passed to downstream equipment may be controlled by reducing the absorption or by removal from the rich amine. In general, the absorption of HC and BTEX from a gas into the rich amine solution can be reduced by higher rich amine temperatures, lower column pressures, lower amine circulation rates, higher acid gas loadings, lower amine concentrations and a lower molecular weight amine. For liquid liquid systems such as treating of LPG s, the above guidelines apply except that higher rich amine temperatures increase the absorption, and the pressure has little effect. The methods for removing the HC and BTEX from the rich amine include a hot flash and a stripping column. In general, the hot flash is not very effective while the stripping column can remove up to 7% or higher. LITERATURE CITED 1. McIntyre, G.D., V.N. Hernandez-Valencia and K.M. Lunsford, Recent GPA Data Improves BTEX Predictions for Amine Sweetening Facilities, Proceedings of the 8 th Gas Processors Association Convention, 21. 2. Valtz, A., P. Guilbot, and D. Richon, Amine BTEX Solubility, GPA Research Report, Project 971, Armines, Paris, France, 22. 3. Lawson J.D. and A.W. Garst, Hydrocarbon Gas Solubility in Sweetening Solutions: Methane and Ethane in Aqueous Monoethanolamine and Diethanolamine, Journal of Chemical and Engineering Data, Vol 21, No 1, Pg 3-32, 1976. 4. Dingman J.C. Don t Blame Hydrocarbon Solubility for Entrainment Problems in Amine Treating Systems, Presented at the 1986 AIChE Meeting, Miami Beach, Florida, 1986. 5. Carroll, J.J., F.-Y. Jou, A.E. Mather, and F.D. Otto, Phase Equilibria in the System Water- Methyldiethanolamine-Propane, AIChE Journal, Vol 38, No 4, Pg 511-52, 1992. 6. Jou, F.-Y., J.J. Carroll, A.E. Mather, and F.D. Otto, Phase Equilibria in the System n- Butane-Water-Methyldiethanolamine, Fluid Phase Equilibria, Vol 116, Pg 47-413, 1996. 7. Jou F.-Y., F. D. Otto, and A.E. Mather, Solubility of Ethane in Aqueous Solutions of Triethanolamine, Journal of Chemical and Engineering Data, Vol 41, No 4, Pg 794-795, 1996. 8. Jou, F.-Y., J.J. Carroll, A.E. Mather, and F.D. Otto, The Solubility of Methane in Aqueous Solutions of 2-(2-Aminoethoxy)ethanol, Ind. Eng. Chem. Res., Vol 37, No 8, Pg 3519-3523, 1998. 15
9. Jou, F.-Y., J.J. Carroll, A.E. Mather, and F.D. Otto, The Solubility of Methane and Ethane in Aqueous Solutions of Methyldiethanolamine, Journal of Chemical and Engineering Data, Vol 43, No 5, Pg 781-784, 1998. 1. Carroll, J.J., F.-Y. Jou, A.E. Mather and F.D. Otto, The Solubility of Methane in Aqueous Solutions of Monoethanolamine, Diethanolamine and Triethanolamine, The Canadian Journal of Chemical Engineering, Vol. 76, Pg 945-951, 1998. 11. Hegarty, M. and D. Hawthorne, Application of BTEX/Amine VLE Data at Hanlon Robb Gas Plant, Proceedings of the 78 th Annual GPA Convention, Pg 128-14, Nashville, Tennessee, 1999. 12. Addicks, J., G. Owren, K. Tangvik, Solubility of carbon dioxide and methane in aqueous alkanolamine solutions, Proceedings of the International Gas Research Conference, 21. 13. Critchfield, J., P. Holub, H.-J. Ng, A.E. Mather, F.-Y. Jou, and T. Bacon, Solubility of Hydrocarbons in Aqueous Solutions of Gas Treating Amines, Proceedings of the Laurance Reid Gas Conditioning Conference, Pg 199-227, Norman, Oklahoma, 21. 14. Jou, F.-Y., H.-J. Ng, J.E. Critchfield, and A.E. Mather, Solubility of propane in aqueous alkanolamine solutions, Fluid Phase Equilibria, Vol. 194-197, Pg 825-83, 22. 15. Addicks, Jan and G.A. Owren, Solubility of Carbon Dioxide and Methane in Aqueous Methyldiethanolamine Solutions, Journal of Chemical Engineering Data, Vol. 47, Pg 855-86, 22. 16. Harbison, J.L. and J.C. Dingman, Mercaptan Removal Experiences in DGA Sweetening of Low Pressure Gas, Proceedings of the Laurence Reid Gas Conditioning Conference, Pg H-1 through H-7, Norman, Oklahoma, 1972. 17. Huval, M., and H. Van de Venne, Gas Sweetening in Saudi Arabia in Large DGA Plants, Proceedings of the Laurance Reid Gas Conditioning Conference, Pg D-1 through D-27, Norman, Oklahoma, 1981. 18. Law, J.A., and R.D. Seidlitz, The Hanlon Robb Project A Modern Canadian Gas Sweetening Plant Comes on Stream, Proceedings of the 63 rd Annual GPA Convention, Pg 1-14, New Orleans, Louisiana, 1984. 19. Morrow, D.C., Removal/Disposal of BTEX Components in Amine Systems, Permian Basin Regional GPA Meeting, 1996. 16
2. Skinner, F.D., D.L. Reif, A.C. Wilson, and J.M. Evans, Absorption of BTEX and Other Organics and Distribution Between Natural Gas Sweetening Unit Streams, SPE 37881 Society of Petroleum Engineers, Presented at 1997 SPE/EPA Exploration and Production Environmental Conference, Dallas, Texas, March 3-5, 1997. 21. Wallender, J.W. and D.C. Morrow, Reduction of BTEX Emissions From Amine Plant Acid Gas Streams, GRI Gas Industry Air Toxics Conference, San Antonio, Texas, May 24-26, 1999. 22. Morrow, D.C., and K.M., Lunsford, Removal and Disposal of BTEX Components from Amine Plant Acid Gas Streams, Proceedings of the 76 th Annual GPA Convention, Pg 171-173, San Antonio, Texas, 1997. 23. Mather, A.E. and K.N. Marsh, Comments on the Paper Hydrocarbon Gas Solubility in Sweetening Solutions: Methane and Ethane in Aqueous Monoethanolamine and Diethanolamine, Journal of Chemical and Engineering Data, Vol 41, No 5, Pg 121, 1996. 17